Article Figures & Data
Figures
- Figure 1.
Layout of the experimental tasks. A, The monetary incentive delay task, in which participants indicate their excitement in playing for tickets toward a monetary reward. B, The calibration phase of the change detection task, where participants saw an array of either two, four, six, or eight squares and indicated whether the final square that appeared (the probe) was of the same or a different color from the square that appeared in the same location in the previously show array. C, The reward phase of the change detection task, where participants saw an array of either two or six, or four or eight, shapes onscreen depending on their performance in the calibration phase of the task. Once again, participants indicated with a keyboard press whether the final square that appeared (the probe) is the same or a different color from the square that appeared in the same location in the array. Here, if they made the correct decision, they would receive 1 point (low effort trial) or 10 points (high effort trial); they would receive 0 points for an incorrect decision.
- Figure 2.
Distribution of A, depressive trait (BDI) proportion scores (score divided by total possible score), B, perceived stress (PSS) scores, C, anhedonia (DARS) scores, and D, mean reward anticipation scores by sex and effort deployment group. BDI, beck depression inventory; PSS, perceived stress scale; DARS, dimensional anhedonia rating scale.
- Figure 3.
Correlations among depressive traits (BDI), perceived stress (PSS), reward anticipation, and short-term memory ability (K estimate). BDI, beck depression inventory; PSS, perceived stress sale; anticipation, mean reward anticipation; K estimate, visual short-term memory ability (K score).
- Figure 4.
Binomial logistic regression predicting whether participants were in the high effort versus low effort group. High effort group,
- Figure 5.
Accuracy and choice latency in the choice phase of the change detection task by motivation group (high effort vs low effort group). The choice latency plot has been log-transformed on the y-axis to more clearly show small choice latency values. High effort group,
- Figure 6.
A, Accuracy and B, choice latency in the reward phase of the change detection task by working memory ability (K estimate). The choice latency plot has been log-transformed on the y-axis to more clearly show small choice latency values.
- Figure 7.
Outputs of a hierarchical Bayesian drift diffusion model fit to all participants in the effort choice task, divided by effort group. High effort participants had a significantly higher A, drift rate, B, starting point and C, boundary separation compared to low effort participants, but did not significantly differ on D, non-decision time compared to low effort participants.
Tables
Group Sex n Mage SDage Minage Maxage MBDIprop SDBDIprop MPSS SDPSS MDARS SDDARS Mant. SDant. High effort Female 263 20.34 2.37 16 42 0.27 0.18 20.83 6.21 35.61 11.71 3.94 1.81 Male 80 20.70 2.34 17 32 0.19 0.14 17.69 5.30 37.69 9.62 4.29 1.75 Other 5 20.40 0.89 19 21 0.17 0.10 19.60 3.65 43.00 10.07 4.80 1.64 Low effort Female 69 20.01 1.94 16 26 0.28 0.16 20.68 6.54 38.04 8.94 4.00 1.79 Male 12 20.92 2.97 18 28 0.19 0.14 21.08 4.38 39.17 8.61 3.96 1.51 BDI prop: beck depression inventory II proportion score (score divided by max score); PSS: perceived stress scale score; DARS: dimensional anhedonia rating scale; Ant.: mean reward anticipation in the behavioral monetary incentive delay task; high effort group: >70% HR trials selected; low effort group: ≤70% HR trials selected.
- Table 2.
Binomial logistic regression predicting whether participants were in the high effort versus low effort group
Binomial logistic regression: high versus low effort group Working memory ability (K estimate) 0.21* (0.09) BDI (prop. score) −0.78 (0.96) PSS 0.00 (0.03) Reward anticipation 0.00 (0.07) Num. Obs. 429 AIC 419.7 BIC 440.0 Log. Lik. −204.833 RMSE 0.39 High effort group,
+p < 0.1, *p < 0.05, **p < 0.01, ***p < 0.001.
Linear model: proportion of high effort trials selected (Intercept) 0.84*** (0.04) Working memory ability (K estimate) 0.02** (0.01) BDI (prop. score) −0.03 (0.07) PSS 0.00 (0.00) Reward anticipation 0.00 (0.00) Num. Obs. 429 R2 0.018 R2 Adj. 0.009 AIC −274.2 BIC −249.8 Log. Lik. 143.105 RMSE 0.17 High effort group,
+p < 0.1, *p < 0.05, **p < 0.01, ***p < 0.001.
Group Effort level MAccuracy SDAccuracy MLatency SDLatency High effort High 0.70 0.11 1.27 1.04 High effort Low 0.95 0.11 2.54 1.99 Low effort High 0.70 0.11 1.23 1.10 Low effort Low 0.90 0.11 1.60 1.92 High effort group,
- Table 5.
Multi-level model analysis coefficients and standard errors for accuracy in the reward phase of the task
Multi-level model: accuracy (Intercept) 0.64*** (0.02) Sex 0.01 (0.01) Working memory ability (K estimate) 0.01** (0.00) BDI (prop. score) −0.01 (0.04) PSS 0.00 (0.00) Reward anticipation 0.00 (0.00) Effort level 0.25*** (0.01) SD (intercept participant) 0.04 SD (observations) 0.10 Num. Obs. 542 R2 Marg. 0.592 R2 Cond. 0.656 AIC −840.0 BIC −801.3 ICC 0.2 RMSE 0.09 BDI (prop. score), depression score on the beck depression inventory; PSS, perceived stress scale. Proportion scores are scores divided by total possible score. AIC, Akaike information criterion; BIC, Bayesian information criterion; ICC, intraclass correlation; RMSE, root mean squared error.
+p < 0.1, *p < 0.05, **p < 0.01, ***p < 0.001.
- Table 6.
Multi-level model analysis coefficients and standard errors for choice latency in the reward phase of the task
Multi-level model: choice latency (Intercept) 1.25*** (0.34) Sex −0.07 (0.17) Working memory ability (K estimate) 0.05 (0.05) BDI (prop. score) −1.11* (0.54) PSS 0.01 (0.01) Reward anticipation −0.04 (0.04) Effort level 1.06*** (0.13) SD (intercept participant) 0.18 SD (observations) 1.53 Num. Obs. 542 R2 Marg. 0.114 R2 Cond. 0.127 AIC 2042.9 BIC 2081.6 ICC 0.0 RMSE 1.51 BDI (prop. score), depression score on the beck depression inventory; PSS, perceived stress scale. Proportion scores are scores divided by total possible score. AIC, Akaike information criterion; BIC, Bayesian information criterion; ICC, intraclass correlation; RMSE, root mean squared error.
+p < 0.1, *p < 0.05, **p < 0.01, ***p < 0.001.
In this issue
